How vivasvana1/sparepartsinventory works
Main Components
• Data Ingestion & Storage
• Data Preparation & Feature Engineering
• Forecasting Engine
• Optimization Framework
• Simulation & Validation
• Visualization & Dashboards
• External Services & Tools
Relationships & Data Flow
• Feedback loops where simulation results update models or dashboards.
Architectural Patterns
• Modular pipeline: each stage is a separate “service” or notebook group.
• Hybrid execution: Python notebooks + external solvers (CPLEX) + GUI tools (anyLogistix, Power BI).
• Batch processing: sequential notebook runs.
• File‐based integration: CSV/Excel/OPL files exchanged between components.
Technologies & Frameworks
• Python (pandas, NumPy, statsmodels, fbprophet, scikit-learn)
• Jupyter Notebooks
• PuLP & IBM CPLEX (MILP solver)
• anyLogistix™ (network simulation)
• Power BI
• Conda environment
Capturing Details
• MILP objective & constraints
• Service-level tiers (A/B/C)
• Regional lead‐time multipliers
Final Touches
• Components are clearly labeled.
• Footnote the technologies (.env, Python libs, conda, CPLEX, anyLogistix, Power BI).